Sunday, December 13, 2009

Google demonstrates quantum computer image search

http://www.newscientist.com/article/dn18272-google-demonstrates-quantum-computer-image-search.html

Google's web services may be considered cutting edge, but they run in warehouses filled with conventional computers. Now the search giant has revealed it is investigating the use of quantum computers to run its next generation of faster applications.

Writing on Google's research blog this week, Hartmut Neven, head of its image recognition team, reveals that the Californian firm has for three years been quietly developing a quantum computer that can identify particular objects in a database of stills or video.

Google has been doing this, Neven says, with D-Wave, a Canadian firm that has developed an on-chip array of quantum bits – or qubits – encoded in magnetically coupled superconducting loops.

The team set themselves the challenge of writing an algorithm for the chip that could learn to recognise cars in photos, and reported at the Neural Information Processing Systems conference in Vancouver, Canada, this week that they have succeeded.
Chasing cars

Using 20,000 photographs of street scenes, half of which contained cars and half of which didn't, they trained the algorithm to recognise what cars look like by hand-labelling all the cars with boxes drawn around them.

After that training, the algorithm was set loose on a second set of 20,000 photos, again with half containing cars. It sorted the images with cars from those without faster than an algorithm on a conventional computer could – faster than anything running in a Google data centre today, Neven says.

Classical computers use what is known as a von Neumann architecture, in which data is fetched from memory and processed according to rules defined in a program to generate results that are stored. It is pretty much a sequential process, though multiple versions of it can run in parallel to speed things up a little.

Quantum computers, however, promise much faster processing, by exploiting the principle of quantum superposition: that a particle such as an ion, electron or photon can be in two different states at the same time. While each basic "bit" of data in a conventional computer can be either a 1 or a 0 at any one time, a qubit can be both at once.
Quantum argument

D-Wave's Chimera chip launched to great media interest. But there has been some dispute over whether it is actually a quantum computer, which Neven acknowledges.

"It is not easy to demonstrate that a multi-qubit system such as the D-Wave chip exhibits the desired quantum behaviour, and physicists are still in the process of characterising it," he writes.

Google's quantum move is understandable, says Winfried Hensinger, reader in quantum, atomic and optical physics at the University of Sussex in Brighton, UK.

"Quantum computing has the potential to make search problems much easier to solve – so it is no surprise that Google finds it extremely important to get involved in this emerging area," he says.

"I expect more and more companies to pursue research in quantum computing due to its vast potential not only in search but also for a multiplicity of other problems," he adds.

However, he expects that while questions remain over the exact capabilities of D-Wave's hardware, future developments will centre on different hardware. "It is widely accepted that trapped ions are the most successful implementation of quantum technology."

If you would like to reuse any content from New Scientist, either in print or online, please contact the syndication department first for permission. New Scientist does not own rights to photos, but there are a variety of licensing options available for use of articles and graphics we own the copyright to.